Rapid and accurate acquisition of landslide inventories is essential for effective disaster relief. Deep learning-based pixel-wise semantic segmentation of remote sensing imagery has greatly advanced in landslide mapping. However, the heavy dependance on extensive annotated labels and sensitivity to domain shifts severely constrain the model performance in unseen domains, leading to poor generalization. To address these limitations, we propose LandsDANet, an innovative unsupervised domain adaptation framework for cross-domain landslide identification. Firstly, adversarial learning is employed to reduce the data distribution discrepancies between the source and target domains, thereby achieving output space alignment. The improved SegFormer serves as the segmentation network, incorporating hierarchical Transformer blocks and an attention mechanism to enhance feature representation capabilities. Secondly, to alleviate inter-domain radiometric discrepancies and attain image-level alignment, a Wallis filter is utilized to perform image style transformation. Considering the class imbalance present in the landslide dataset, a Rare Class Sampling strategy is introduced to mitigate bias towards common classes and strengthen the learning of the rare landslide class. Finally, a contrastive loss is adopted to further optimize and enhance the model’s ability to delineate fine-grained class boundaries. The proposed model is validated on the Potsdam and Vaihingen benchmark datasets, followed by validation in two landslide scenarios induced by earthquakes and rainfall to evaluate its adaptability across different disaster domains. Compared to the source-only model, LandsDANet achieved improvements in IoU of 27.04% and 35.73% in two cross-domain landslide disaster recognition tasks, respectively. This performance not only showcases its outstanding capabilities but also underscores its robust potential to meet the demands for rapid response.

Soil organic carbon (SOC) and soil inorganic carbon (SIC) are key components of soil carbon pools in arid ecosystems, playing a crucial role in regional carbon cycling and climate change mitigation. However, the interactions between these two forms of carbon in arid alpine ecosystems remain underexplored. This study was conducted in the Heihe River Basin (HRB) in the northeastern Qinghai–Tibet Plateau, focusing on the distribution and dynamics of SOC and SIC in deep soil layers. Using data from 329 samples collected from 49 soil profiles extending to the bedrock, combined with path analysis, we explored the inter-relationships between SOC and SIC and quantified the influence of environmental factors. The results showed that (1) SOC exhibited a unimodal distribution with elevation, peaking at 3300–3600 m, while SIC continuously decreased with elevation, with reduction rates ranging from −0.39% to −31.18%; (2) SOC and SIC were significantly positively correlated (r = 0.55, p < 0.01), with SOC decreasing with depth and SIC showing an inflection point at 50 cm depth; (3) SOC was primarily driven by nutrient factors, such as total nitrogen (TN), with a path coefficient of 0.988, while SIC was influenced by abiotic factors, including potential evapotranspiration (PET), with a coefficient of −1.987; (4) SOC density accounted for 81.62% of the total soil carbon pool, playing a dominant role in carbon storage, whereas SIC density exhibited dynamic changes, particularly at depths of 110–150 cm. These findings advance our understanding of deep soil carbon dynamics in arid alpine ecosystems and provide critical data for improving carbon management strategies in similar regions.